1. "Do we really know what makes us healthy" by Gary Taubes published in the New York Times, September 16, 2007. The summary below comes from Peter Hooper's Stat 141 website.

The article discusses why causal effects suggested in observational studies are often not confirmed or are even contradicted in subsequent clinical trials.  A main focus of the article is Hormone Replacement Therapy (HRT). Reports in 1985 from the Nurse's Health Study (a large prospective observational study) suggested that HRT lowers the risk of heart attacks. A subsequent clinical trial found that HRT actually increases the risk of heart attacks among older women. The author discusses the difficulties in drawing valid conclusions about cause-and-effect relationships from observational studies and recommends skepticism when reading reports from such studies.

2. "UBC professors question effectiveness of Cold-FX" published by Calgary Herald, February 25, 2006.

In 2006 Cold-fX ranked as the  country's bestselling cold and flu remedy. In fact, many people claim that Cold-fX supposedly cleared their symptoms within 3-4 days. How do they know ? How can they say for sure that their cold symptoms were moderated because of that product? Maybe their symptoms would have cleared anyway as colds just come and go?

Two pharmaceutical researchers from UBC with a background in testing drug companies' claims tested the claims of the Cold-fX maker CV Technologies and concluded that there is no evidence that Cold-fX statistically reduced the percentage of subjects who ended up getting a cold. In the interesting article above their claims are countered by the company's president. The article is a good introduction to experiment design in undergraduate courses.

In 2007 Health Canada approved the CV Technologies claims that the product "helps reduce the frequency, severity and duration of cold and flu symptoms by boosting the immune system."

 
Statistical Laboratories Department of Mathematical and Statistical Sciences, University of Alberta